Ataxia-telangiectasia (AT) is an autosomal recessive disorder characterized by cerebellar degeneration and oculocutaneous telangiectasia, accompanied by immunodeficiency, infertility, growth retardation, sensitivity to the effects of ionizing radiation and cancer predisposition. Heterozygous carriers may have an increased predisposition to cancer, particularly breast cancer. To address the complex relationship between gene function and the pleiotropic AT phenotype, a murine model of ataxia-telangiectasia was created by disrupting the Atm locus via gene targeting. Mice homozygous for the disrupted Atm allele recapitulate the AT phenotype in humans, providing a mammalian model in which to study the pathophysiology of this disorder. We have used genetic and biochemical approaches to try to understand the mechanisms by which a single gene disruption can result in such a pleiotropic phenotype, and to understand pathways regulated by Atm. Atm-deficient mice are completely infertile. Male and female gametogenesis is severely disrupted in Atm-deficient mice as early as leptonema of prophase I, resulting in apoptotic degeneration. Rad51 foci are not assembled properly on unpaired axial elements in leptonema of mutant spermatocytes, and p53, p21 and Bax are elevated in these mutant testes. In Atm/p53 or Atm/p21 double mutants, spermatogenesis progresses further into pachytene stages, but not to diplonema, although Rad51 assembly remains defective. Thus, Atm is absolutely required for Rad51 assembly onto the axial elements, and for suppressing p53, p21 and Bax levels, suggesting that Atm participates in the regulation or surveillance of meiotic recombination and progression. In response to ionizing radiation (IR), Atm is part of a DNA damage- response pathway that involves p53. p53 is a multifunctional protein that simultaneously regulates distinct downstream pathways controlling cell cycle progression and apoptosis. However, the mechanisms by which p53 differentially activates downstream pathways are unknown. We have investigated this problem in Atm-deficient mice. Our data demonstrate that after IR, Atm acts via p53 to activate cell cycle checkpoint pathways and induce p21. However, p53-dependent or p53-independent apoptotic responses and Bax induction are independent of Atm function. IR-induced thymic apoptosis was suppressed in Atm/p53 double mutant mice, but not in Atm/p21 double mutants, demonstrating that these IR-mediated apoptotic responses are p53- dependent. These results support a model in which upstream effectors such as Atm selectively activate p53 to regulate specific downstream pathways, providing a mechanism for controlling distinct cell cycle and apoptotic responses. Finally, we have demonstrated that c-abl is directly phosphorylated by ATM, identifying the first direct target of ATM, but the biological significance of this is unknown. In summary, we have discovered tissue-specific differences in the biochemical responses to loss-of-function of Atm, and provided evidence that distinct Atm-dependent pathways in different tissues may be responsible for the pleiotropic phenotype of Atm- deficient mice.